Application of a simple GPC with robust behaviour, to a sugar factory

Abstract

This paper presents the application of a Generalized Predictive Controller (GPC) to the diffusion process in a sugar factory. The controIler was embedded in the existing control system, needing the same computational requirements as a PID routine. The control law is extremely simple to compute and the tuning is straightforward since a low order model is used. Due to model uncertainties that. appear , when working at different operating points, the original GPC algorithm is improved by the use of the so-called T polynornial, which increases the stability robustness by filtering the predictions.